Skip to main content

Reply to this post | See parent post | Go Back
View Post [edit]

Poster: xtifr Date: Aug 29, 2008 2:53pm
Forum: etree Subject: Flac holding its own

I'm interested to see the link, but my money's on FLAC. I know some of the people who were involved with FLAC development, and their work is solid, and they say it should work, and from what I know (a fair amount) about Ogg, there's absolutely no reason why the wrapper should make an appreciable difference in any cases.

I'm currently 6:30 into a 33 minute FLAC from here at the Archive. I'll report back when it either finishes or flakes out. But my money is on the former.

Later: ok, I've just spent an enjoyable half-hour-plus listening to an entertaining medley: http://www.archive.org/download/rad2000-04-02.flac16/rad2000-04-02d1t03.flac is 175MB. Is that big enough?

I had absolutely no problems with playback at all. At the 32 minute mark, I paused, because resuming from a pause is something that will break with formats not designed for web streaming. When I resumed, the music started right back up (and my network activity monitor lit up again, showing that the music was actually streaming, rather than being held in a local cache or something).

How big was our wager? Because I'm pretty sure I just won. Pay me when you can. :)

At this point, I'd have to say that all the available evidence suggests that the problems you had streaming FLAC "a few moons back" were in your setup or the software you used, rather than anything inherent with FLAC, which should and does stream just fine.

This post was modified by xtifr on 2008-08-29 21:53:44

Reply to this post
Reply [edit]

Poster: Samizdat Date: Sep 8, 2008 8:10am
Forum: etree Subject: Tutorial Amended -- 2008-09-08 Update == Brahms Show Fixed ==

2008-09-08 -- ==Brahms Show Fixed ==

One of the premises of the parent post, which is that the ogg-FLAC file/format is the way to "stream the lossless files" (from the Internet Archive), is evidently mistaken. As far as I now understand the ogg-FLAC file/format for Web-streaming purposes, some players in some situations do not support FLAC but do support ogg-FLAC. The converse is probably also true, as ogg-FLAC is still relatively new.

This post was modified by Samizdat on 2008-09-08 15:10:33

Reply to this post
Reply [edit]

Poster: xtifr Date: Sep 8, 2008 2:01pm
Forum: etree Subject: Re: Tutorial Amended -- 2008-09-08 Update == Brahms Show Fixed ==

Nothing wrong with your tutorial that I can see--but I don't believe that Ogg-Flac is intended as "the" way to stream lossless audio. From what I've been able to determine, the point of Ogg-Flac is not to be a pure audio format, but to allow the use of Flac in multimedia presentations, i.e. (mainly) coupled with Theora. Theora is normally coupled with Vorbis in an Ogg container, but Theora+Flac could be really nice for concert videos. Maybe with a separate channel using Speex for 'directors commentary" or something similar. :)

Reply to this post
Reply [edit]

Poster: Samizdat Date: Sep 15, 2008 10:46am
Forum: etree Subject: Re: Tutorial Amended -- 2008-09-08 Update == Brahms Show Fixed ==

Thanks for holding my feet to the fire, xtifr. It's more important to get it right than be right. (As it turns out, [in Linux Ubuntu, anyway] the original Brahms 48 KHz FLAC file Web-streams perfectly in a nifty little player called Audacious). That's 49:39 of playback in a single 240 MB FLAC file -- impressive. Unfortunately, Audacious is one of those players which doesn't support ogg-FLAC -- at least not out-of-the-box.

Expanding on your mention of Theora -- using Linux Ubuntu and the screencasting app recordMyDesktop, I can combine video and audio, writing the output as Ogg Theora with a .ogg extension. Whether recordMyDesktop can combine lossless audio and video or not I don't know. Considering for instance the stealthy MPEG audio Bartok Radio Budapest uses occasionally, sniffing out lossy-sourced audio is hard enough. I can only imagine what a limited PC and some patchwork processing tools can do video-wise, in ferreting out lossy-sourced material.

This post was modified by Samizdat on 2008-09-15 17:46:15

Reply to this post
Reply [edit]

Poster: Arrond_Wolf Date: Oct 20, 2008 12:13pm
Forum: etree Subject: Re: Tutorial Amended -- 2008-09-08 Update == Brahms Show Fixed ==

This is a fairly absurd discourse.
Hypothetically,
1) a large enough out-going bandwidth from the files source 2) a large enough in-coming bandwidth to allow for real-time processing of a FLAC codec
you can stream any type or size media file with a media player that supports streaming FLAC...not ogg-FLAC.

1st- I tried to make an *.m3u by adding the url for a FLAC file in a new playlist in Winamp.
No Go.
2nd try- Foobar2000 v.0.9.5.3. Created a new playlist using FLAC download links from LMA. Success.

As soon as my current file transfers finish, I'll upload an *.m3u playlist that contains links to set 1 FLACs from http://www.archive.org/details/um2008-10-17.mk012.flac16.

here's the link:http://www.1424multimediastuidos.com/Music/Umphreys_McGee/um2008-10-17.mk012.set1.m3u

It worked when i made a new playlist and open the location as the url above and it streams fines with zero lag.

Keep It Greasy,
Paul Staats

Reply to this post
Reply [edit]

Poster: Samizdat Date: Sep 8, 2008 8:39am
Forum: etree Subject: == 2008-09-08 == Brahms Show Fixed == *Still* Stand by My Tutorial

2008-09-08 - Brahms Show Fixed

I still stand by my tutorial -- been busy, though, trying to figure out odd FLAC file behavior. The Euroradio concert herein linked had a sample rate of 48 KHz, with which several media players I tried had no idea what to do. I resampled the 48 KHz FLAC file to 44.1 KHz, which seemed to solve the compatibility problem, and in addition rendered a truer representation of the performance. Mind you, I'm no music theorist, but to my ear, the 48 KHz version is a bit jaunty for a work written in a minor key:*

Brahms - Tragic Overture in D Minor, Op. 81 (plus Stravinsky's Petrushka and a piece by Mendelssohn).

* The original 48 KHz file temporarily remains in the item folder, but as it's strictly a test file, there is no corresponding checksum file.

This post was modified by Samizdat on 2008-09-08 15:39:41